Guppy colour learning project site GitHub Analysis 2 Analysis 1 Protocols Home

Overview

This page contains detailed protocols for the experiments in ‘Colour biases in learned foraging preferences in Trinidadian guppies’. We provide step by step instructions on how to perform the training, testing, and behavioural quantification described in the manuscript. If you would like to see details on housing and husbandry conditions during the experiment please refer to section 2 of the manuscript.

For this project, we wanted to determine whether object preferences were plastic for both green and blue objects in Trinidadian guppies. Given these colours differ in their foraging relevancy for guppies we wanted to determine whether the shift in preference would be unconstrained. Our protocol aimed to manipulate object preferences for the green and blue objects (Figure 2) of individual guppies by manipulating their foraging experiences with these objects. This was done by providing food reinforcement for approaching one of the two colours.

On the first trial, guppies were placed in the test tank with two objects, one blue and one green, but with empty food strips to measure initial preferences. Training then occurred once daily between 10h00 to 16h00 for 20 days constituting 20 trials. During training, the location of the rewarding object (left or right) was randomized across days and individuals. That is, only the object’s appearance provided a reliable cue to food location. On the 21st day, an unreinforced test was given where new duplicates of the objects were presented with empty plastic strips attached to them. In both experiments this unreinforced trial (Trial 21, Probe 1) was used to assess whether learning had occurred. At this point, experiment 1 ended. In experiment 2 an additional four test trials were conducted to see how robust learning was in both groups across different contexts.

An experimental timeline representing the order of the trials. Experiment 1 ends at trial 21. Experiment 2 ends at trial 29.

Figure 1: An experimental timeline representing the order of the trials. Experiment 1 ends at trial 21. Experiment 2 ends at trial 29.

Sample footage

In the video below we provide sample footage of the behavioural assays conducted during the experiments. The footage below is of a green-trained guppy throughout experiment 2 and demonstrates the types of trials guppies were exposed to in this project (see Figure 2 for details on trials).

Test trials

For the initial test and final tests guppies were presented with the green and blue objects with empty food strips attached to them. You can see an example of this above. The following protocol was used for testing:

  1. Fill the test tank with water from several home tanks to the 8 cm line.
  2. Transfer the guppy from the home tank to the test tank via a white opaque plastic cup where the green and blue objects had already been placed with new and blank plastic food strips.
  3. Allow guppies to acclimate in a 20 cm tall, open-ended glass holding cylinder (7 cm diameter) after being transferred into the test tank for 120 seconds.
  4. After acclimation lift the release glass was lifted and give guppies 5 minutes to move about the tank and feed from the objects. At this point the experimenter leaves the room.
  5. After 5 minutes the guppy is transferred back to the home tank via the white opaque plastic cup.
  6. Repeat steps 2 to 5 for all guppies.

Training and refresher trials

Guppies were rewarded for foraging from either the blue object or the green object, forming the two experimental treatments, blue-rewarded and green-rewarded. The food reward was a gelatin mixture of flake food which was placed on the back of the objects facing away from the guppy so that the rewarding object had to be approached and investigated for the food reward to be discovered. To see how to make the food reward go to Food strips. The protocol was as follows:

  1. Fill the test tank with water from several home tanks to the 8 cm line.
  2. Transfer the guppy from the home tank to the test tank via a white opaque plastic cup where the green and blue objects have already been placed into the test tank. The objects should be placed on their designated velcro piece which are each 6 cm from the side of the tank with one on the left and right end of the tank. The two objects are 38 cm apart.
  3. Dip the food strip in some water to remove excess flakes. Keep dipping until the strip does not shed flakes upon being dipped in water (about 8 to 10 dips)
  4. Attach a food strip to the back of the correct rewarding object for that trial using blunt tweezers. Consult the data sheet to confirm which guppy is rewarded to which object and which side of the tank the rewarding object should be placed.
  5. Gently pour the guppy into a 20 cm tall, open-ended glass holding cylinder (7 cm diameter). Allow the guppy to acclimate in the holding cylinder for 120 seconds.
  6. After acclimation the release glass should be slowly lifted and the guppies given 5 minutes to move about the tank and feed from the objects. Set a timer for 5 minutes and 20 seconds and leave the room once you have lifted the cup. Come back after 5 minutes and 20 seconds.
  7. Stop the video recording using the remote.
  8. Transfer the guppy back to the home tank via the white opaque plastic cup.
  9. Repeat steps 3 to 8 for all guppies.

Food strips

The food strips that were attached to the objects were made with the following protocol:

  1. Pour 50 mL of conditioned water into a 100 mL beaker.
  2. Add a serving of decapsulated brine shrimp egg to the water and stir then add a pinch of Tetramin flake food and stir.
  3. Add a packet of Knox gelatin to the water and stir for about 15 seconds or until the powder fully dissolves, then wait another 15 seconds.
  4. Use a transfer pipette to transfer the food mixture from the beaker onto 1.5 cm x 0.5 cm (w x h) strips of plastic.
  5. Spread another pinch of flake food on top of the food mixture on the plastic strips while it is still wet.
  6. Wait one hour for food to fix onto the plastic strips.

Objects

(A) The two objects used in experiment 1. Blue-trained guppies were trained to the left object and green-trained guppies were trained to the right object. The manufacturer’s colour name for the blue object is ‘dark azur (hex #078BC9) and for the green object is ‘bright yellowish green’ (hex # BBE90B). (B) The 7 pairs of objects in experiment 2. Object pairs 1,2,3,4, and A had the same colour properties as the object pair used in experiment 1. The bricks for the generalization objects were 'dark green' (hex #237841) and 'bright blue' (#0055BF) as specified by the manufacturer. The numbered object pairs are used for training while the lettered object pairs are used for testing. All Lego objects were placed in the tank with the studs facing up except object pair 1 which was placed in the tank with studs facing to the side as in the image. The experimental timeline represents the order of the trials. Experiment 1 ends at trial 21. Experiment 2 ends at trial 29.

Figure 2: (A) The two objects used in experiment 1. Blue-trained guppies were trained to the left object and green-trained guppies were trained to the right object. The manufacturer’s colour name for the blue object is ‘dark azur (hex #078BC9) and for the green object is ‘bright yellowish green’ (hex # BBE90B). (B) The 7 pairs of objects in experiment 2. Object pairs 1,2,3,4, and A had the same colour properties as the object pair used in experiment 1. The bricks for the generalization objects were ‘dark green’ (hex #237841) and ‘bright blue’ (#0055BF) as specified by the manufacturer. The numbered object pairs are used for training while the lettered object pairs are used for testing. All Lego objects were placed in the tank with the studs facing up except object pair 1 which was placed in the tank with studs facing to the side as in the image. The experimental timeline represents the order of the trials. Experiment 1 ends at trial 21. Experiment 2 ends at trial 29.


Materials information

Colours

The colours used in the experiment are listed in Table 1.

Table 1: Manufacturer colour details for the Lego brick colours used in the experiment. Bright yellowish green and dark Azur were the green and blue used for all bricks except the generalization objects. The bricks for the generalization objects were dark green and bright blue as specified by the manufacturer.
Object Colour Colour Family Exact Colour Hex Value
Green Dark Green Bright Yellowish Green BBE90B
Green Dark Green Dark Green 237841
Blue Blue Dark Azur 078BC9
Blue Blue Bright blue 0055BF

Bricks

The objects were 6 pairs of Lego blocks with the manufacturer colours A and B. The specific models for the object pairs are listed in Table 2. As of 2021, these bricks can be looked up and/or purchased using the official Lego store’s Choose a brick service by entering the design ID in the Enter an item number or design ID prompt of the Enter set number section.

Table 2: The specific models for the Lego object pairs used in the experiment. Pair 1 was constructed by combining two 1X2 bricks and laying the composite brick on its side.
Object Pair Brick Style Design ID
Pair A BRICK 1X2 3004
Pair B BRICK 2x4 3001
Pair 1 BRICK 1X2 (x2) 3004
Pair 2 BRICK 2X4 3001
Pair 3 & Exp. 1 Green object BRICK 2X2 30165
Pair 4 BRICK 1X1 3005
Exp. 1 Blue object BRICK 1X3 3622

EthoVision behavioural quantification

We quantified behaviour throughout this project using EthoVision XT 11.5. Below are the instructions that were given to an undergraduate student (Mina Araya-Yohannes) tasked with conducting experimental trials, and collecting as well as quantifying behavioural data for a large portion of experiment 2 trials. While EthoVision is proprietary software, we reproduce our protocol for using it in this experiment for those who may have it or intend to use it.

  1. Carefully remove the SD card from the camera and place it into the card reader on the wet lab computer.
  2. Navigate to the video files which are located in L://MP_ROOT/
  3. Copy the 16 .mp4 video files from the trials of the current day into a folder named after the trial number (e.g., trial-x/) found within the colour-learning-experiment-2/videos/ folder.
  4. Rename the .mp4 files based on the IDs of the fish. Each video file should be named after a specific fish.
  5. Open EthoVision XT 11.5
  6. Under New experiment select New from template
  7. Select Use a custom template
  8. Select Browse... and navigate to colour-learning-experiment-2/trial-1/trial-1.evxt to use the experiment template from this file
  9. Click OK
  10. Click Browse... and navigate to colour-learning-experiment-2/ and name the file trial-x where x is the number of the trial for that day (this can be found on the data sheet).
  11. Click the Arena Settings tab and select Arena Settings 1
  12. Go to the panel on the right and right click Background.
  13. Navigate to the first video file for that day and click Open.
  14. Click Grab and make sure the arena graphics still line up then click Validate Setup.
  15. Click Trial List
  16. Click Add Trials in the top left corner and add 16 trials
  17. Click Acquisition (0 acquired)
  18. Scroll through the video footage until the point where the release cylinder is removed from the tank
  19. Wait 20 seconds to allow the water to calm
  20. Click the small camera icon on the left of the Playback Control to set the background image.
  21. Click the red circle to start processing the video.
  22. Watch for any errors in tracking and make note of when these occur
  23. Record whether the guppy eats from the rewarded object during the trial on the data sheet
  24. Click ... next to Video under the Settings tab to the right
  25. Select the next video
  26. Repeat steps 18-25 for all videos
  27. Go to the trial list and fill in the variables
  28. Perform track error correction
  29. Under the Analysis tab go to the Results tab and click Statistics & Charts
  30. Click Calculate
  31. Under the Export tab click Statistics
  32. Click OK, the default settings should be sufficient
  33. Navigate to colour-learning-experiment-2/trial-x/Export Files/
  34. Open Google chrome and click the third bookmark tab which contains the Google drive folder named experiment-2-raw-data/
  35. Drag the raw data file Statistics-trial-x into the Google drive folder.
  36. Save EthoVision and close it.